Treating comorbid Depression During Care Transitions Using Relational Agents Project Summary Comorbid depression among individuals with chronic illness such as heart failure (CHF) or chronic pulmonary disease (COPD) is a strong predictor of poor clinical outcomes and high rates of avoidable 30-day readmissions. While depression is a treatable condition, less than 36% of people with depression receive effective treatment. The proposed study aims to design and develop a technology-driven relational agent system to deliver a 5 module, blended approach of cognitive behavioral therapy and self-management education (RA-CBT) to support patients with CHF or COPD and comorbid depression at the time of discharge from an index hospitalization. The relational agent is a human animation program that interacts with patients, integrates best practices from provider-patient communication theory, emulating the face-to-face conversational behavior of an empathic provider emphasizing nonverbal communicative behavior such as gaze, posture, gestures, etc. The RA-CBT program will be accessed via internet by participants during the post-hospital discharge transition period from hospital to home. We will pilot test the feasibility and acceptability of using the RA-CBT system and its effectiveness in treating comorbid depression and reducing 30-day readmissions compared with usual care in 100 chronically ill patients from an urban academic safety net medical center. If successful, this new approach to depression treatment would immediately expand access and scalability for post-discharge mental health support in the care transition.